PM motor voltage feedback(FOC Drive)

[omid2006]

Hi Guys, I'm working on a EV motor control project, the motor type is IPMSM, on a traditional Drive we have found that it measures the motor voltage feedbacks( middle point of inverter legs) by two methods:
1.Analog method: through resistive voltage divider +LPF+ ADC( MCU input);
2.Digital method: opamp (non-inverting) + Digital MCU input;

the inverter is two-level, so we have a pwm type waveform.
My question is that what is the application of those measurements? Does anyone have similar experience?

this Drive uses single shunt current sensing method + Hall position sensor for closed loop

the circuit schematics for both methods are attached.

 

[cupoftea]

Its needed because the motor may suddenly go locked rotor and they need to sense the collapse of the back EMF. Or else it will be for FOC control, whereby they make the motor have max torque at zero speed.

 

[dick_freebird]

Using armature voltage for speed control, you ought to add in the I*R term
early so that speed (not speed-sag) is the fed-back quantity. Not just as a
protection term, but to make the control loop a lot less challenged by
externalities.

 

[omid2006]

Using armature voltage for speed control, you ought to add in the I*R term
early so that speed (not speed-sag) is the fed-back quantity. Not just as a
protection term, but to make the control loop a lot less challenged by
externalities.

thanks for your reply, Can you please explain more?
I think you mean it's application is for sensoreless control yeah?

 

[dick_freebird]

Not entirely "sensor-less". There is still a current sense resistor, its
voltage is scaled (by armature best-guess resistance), inverted and
summed into the armature reading to get back to back-EMF (the
terminal voltage being back-EMF plus I*R).

But no fancy Hall sensors with their amplifier delay, etc.

 

[omid2006]

Not entirely "sensor-less". There is still a current sense resistor, its
voltage is scaled (by armature best-guess resistance), inverted and
summed into the armature reading to get back to back-EMF (the
terminal voltage being back-EMF plus I*R).

But no fancy Hall sensors with their amplifier delay, etc.

For reading with resistive devide maybe yes. but for reading with opamps, definately it usses that directly in control algorithm. because by cutting the one of the opamp inputs, the motor stops and PWM disturbes. I will attach the opamp output here.